bioRxiv preprint doi: https://doi.org/10.1101/2020.08.11.245837; this version posted January 30, 2021. The copyright holder for this preprint (which was not certified by peer review) is the author/funder, who has granted bioRxiv a license to display the preprint in perpetuity. It is made available under aCC-BY 4.0 International license.

1 Red deer Cervus elaphus blink more in larger groups 2 3 Zeke W. Rowe*, Joseph H. Robins* and Sean A. Rands† 4 5 School of Biological Sciences, University of Bristol, Life Sciences Building, Tyndall Avenue, 6 Bristol BS8 1TQ, United Kingdom 7 8 ORCID: ZWR: 0000-0002-9617-5539 9 JHR: 0000-0001-8980-8150 10 SAR: 0000-0002-7400-005X 11 12 * JR and ZR are considered as joint first authors 13 14 † author for correspondence: [email protected] 15 16 17 ABSTRACT 18 19 Most animals need to spend time being vigilant for predators, at the expense of other 20 activities such as foraging. Group-living animals can benefit from the shared vigilance effort 21 of other group members, with individuals reducing personal vigilance effort as group size 22 increases. Behaviours like active scanning or head lifting are usually used to quantify 23 vigilance, but may not be accurate measures as the individual could be conducting them for 24 other purposes. We suggest that measuring an animal’s blinking rate gives a meaningful 25 measure of vigilance: increased blinking implies reduced vigilance, as the animal cannot 26 detect predators when its eyes are closed. We demonstrate that as group size increases in 27 red deer, individuals increase their blink rate, confirming the prediction that vigilance should 28 decrease. Blinking is a simple non-invasive measure, and offers a useful metric for 29 assessing the welfare of animals experiencing an increase in perceived predation risk or 30 other stressors. 31

1 bioRxiv preprint doi: https://doi.org/10.1101/2020.08.11.245837; this version posted January 30, 2021. The copyright holder for this preprint (which was not certified by peer review) is the author/funder, who has granted bioRxiv a license to display the preprint in perpetuity. It is made available under aCC-BY 4.0 International license.

32 INTRODUCTION 33 34 Most animal species spend some part of their lives aggregated together in groups, and 35 many benefits have been proposed and tested for this behaviour [1,2]. For prey species, 36 grouping behaviour can offer protection from predators through both the dilution of individual 37 risk if an attack occurs [3–5] and an increase in the chance of successfully detecting an 38 approaching predator due to the combined vigilance effort of the group [5–7], along with 39 other anti-predator advantages of grouping behaviour such as synchronising activity to dilute 40 risk [8–11]. If an animal is being actively vigilant, it may be unable to conduct (or less 41 efficient at) other important behaviours (like foraging or resting) at the same time (e.g. [12]). 42 Group membership means that vigilance can be pooled among the group members, which 43 could mean that each individual can spend less time being vigilant and more time 44 conducting other fitness-enhancing behaviours. A rich body of theory and research has 45 explored how group size and individual vigilance effort are related [13–16], with much of it 46 focussing on the prediction that individual vigilance effort will decrease as the group 47 becomes larger. This prediction requires each individual to show a trade-off between 48 vigilance and other behaviours, where being actively vigilant either cannot occur at the same 49 time as other behaviours, or leads to a reduction in the efficiency of other behaviours that 50 are conducted at the same time as being vigilant. 51 52 Vigilance is usually assumed to be occurring when an animal is actively scanning its 53 surrounding environment with its head upwards, although there is no obvious consensus in 54 how vigilance is defined in any particular species (see [17] for discussion of this problem in 55 studies on primates). Although scanning behaviour is likely to stop an animal from actively 56 collecting food, this head-up activity may not completely interfere with simultaneous 57 conducted behaviours, such as chewing or social interaction. If a behaviour that is recorded 58 as vigilance allows an individual to do other things at the same time, then we may be falsely 59 assuming that this behaviour incurs the time and attention costs that are associated with 60 vigilance [18]. Without careful experimentation, it is difficult to assess how much of an 61 individual’s attention is devoted to vigilance when we observe scanning or other forms of 62 vigilance-like behaviour, which may add to the huge variation (e.g. [13]) in whether a study 63 demonstrates that individual vigilance is related to group size or not. 64 65 Although it is difficult to define exactly when an individual is being vigilant, we may instead 66 be able to define when it is not able to be vigilant. Blinking (the temporary closure of both 67 eyes, involving movements of the eyelids [19]) is a good example of an activity where an 68 individual is momentarily unable to visually scan the environment. It is an essential

2 bioRxiv preprint doi: https://doi.org/10.1101/2020.08.11.245837; this version posted January 30, 2021. The copyright holder for this preprint (which was not certified by peer review) is the author/funder, who has granted bioRxiv a license to display the preprint in perpetuity. It is made available under aCC-BY 4.0 International license.

69 maintenance behaviour to keep the eyes moist and clean [20], and is conducted tens of 70 times every minute in some species of diurnal mammals [21–23] and birds [24]. Although a 71 blink takes only a fraction of a second, the sum of this loss of visual information over multiple 72 blinks could be substantial for the individual. In humans, spontaneous blinking is 73 accompanied by attentional suppression, where the individual experiences a blackout in 74 visual attention for the duration of the blink, meaning that there is no awareness of the 75 temporary blindness and lack of visual information whilst the blinking is occurring [25,26]. 76 Blinking suppresses activity in both the visual cortex and other areas of the brain that are 77 associated with awareness of environmental change [27]. If we assume that other animals 78 show similar attentional suppression, then they are essentially blind and unaware of 79 changes in their visual environment during each blink, which in turn means that they cannot 80 be vigilant for predators. 81 82 An individual’s blink rate therefore presents a trade-off between the physiological benefits of 83 blinking and the loss of visual information during the blink [20]. If an animal needs to 84 dedicate more time to vigilance in a risky environment, then it has to reduce or suppress 85 blinking to accommodate this increased vigilance. This is anecdotally demonstrated in 86 American crows Corvus brachyrhynchos, which reduce their blink rates when looking at 87 potentially dangerous stimuli [28], and in horses Equus caballus, which decrease their 88 spontaneous blink rate in response to stress-inducing stimuli [29]. This link between blink 89 rate and vigilance implies that blink rate will also be related to group size. As group size 90 increases, theory predicts that individual vigilance can be reduced [5], and so any 91 requirement to suppress blinking will be relaxed. Blink rate may therefore increase with an 92 increase in group size. Evidence supporting this is anecdotal: a comparison of chickens 93 Gallus gallus feeding solitarily or in pairs showed a non-significant increase in blink rate in 94 the group-feeding birds [30], while a comparison of the blink rates of olive baboons Papio 95 anubis [31] showed individuals in a small group blinked less than those in a large group 96 (although the two groups were studied in different years). Here, we test this hypothesis by 97 observing the blink rates of group-living red deer Cervus elaphus. Red deer are a prey 98 species that spend most of their lives in dispersed groups, and females are known to reduce 99 vigilance behaviour as group size increases [32,33]. Because they increase vigilance in 100 smaller groups, we therefore predict that they should also reduce their blink rate to avoid 101 losing visual information about their environment. Given that vigilance has been shown to be 102 related to the sex and age of an individual [32,33], we included these individual 103 characteristics in our analysis. 104 105

3 bioRxiv preprint doi: https://doi.org/10.1101/2020.08.11.245837; this version posted January 30, 2021. The copyright holder for this preprint (which was not certified by peer review) is the author/funder, who has granted bioRxiv a license to display the preprint in perpetuity. It is made available under aCC-BY 4.0 International license.

106 METHODS 107 108 Study area, time and subjects 109 This observational study was conducted on the herd of red deer within the 40.5 hectare deer 110 park in Ashton Court Estate, Bristol, England (51.4440° N, 2.6378° W), which is composed 111 mainly of open grassland, with scattered forestry. The herd, managed by Bristol City 112 Council, consists of c. 110 individuals of varying age and sex, who appear to mix freely. The 113 enclosure is open to the public outside of the rutting season, so the deer are habituated to 114 both dogs and humans. Our observations were conducted over five days during the rutting 115 season; observations were restricted between 1200-1630 h so they were outside of the 116 dawn and dusk peaks of regular rutting activity [34]. 117 118 Behavioural sampling and observations 119 120 A random individual was selected as described in previous research on this herd [35]. The 121 selected individual was observed and dichotomously aged (mature or young) and sexed 122 (male or female). The individual was sexed by the presence of antlers, as after one year of 123 age only males have antlers [36]. Age was identified by an individual's size and morphology 124 (larger individuals were older). If the individual was observed suckling, it was discarded and 125 randomisation was repeated, as young individuals are hard to sex and exhibit behaviours 126 uncommon to the rest of the herd [34]. A count of the total number of young/mature males 127 and females, along with suckling young, was made on three different days. The rounded 128 averages of these five demographic classes were calculated, and used in the 129 pseudoreplication analyses presented (see the Supplementary Information). 130 131 Prior to the study, the observers (ZWR and JHR, who both conducted the measurements 132 described) were trained in identifying the recorded behaviours, and pilot trials ensured 133 repeatability of measurements. A blink was defined as a rapid full closure of the eye. Group 134 size was arbitrarily defined as the number of individuals aggregated no more than five body 135 lengths away from at least one member of the group containing the focal deer, meaning a 136 measured group was composed of individuals associated by a chain rule of association (see 137 [37] for discussion of defining groups by associated neighbours within arbitrary distances). 138 Before starting any set of observations, the observers waited 10 minutes at the site to 139 habituate the deer. Observations were conducted approximately 10-100 m away from focal 140 individuals using a 30× zoom spotting scope. An observation for a single selected individual 141 was recorded for a maximum of 10 minutes. At the start of each minute the group size was 142 counted by one observer, with blink rate (blinks per minute) being continuously recorded for

4 bioRxiv preprint doi: https://doi.org/10.1101/2020.08.11.245837; this version posted January 30, 2021. The copyright holder for this preprint (which was not certified by peer review) is the author/funder, who has granted bioRxiv a license to display the preprint in perpetuity. It is made available under aCC-BY 4.0 International license.

143 each minute by the other observer. If the deer’s eye(s) were obstructed or there was a 144 human/animal disturbance the observations were stopped with the current minute’s 145 measures being disregarded. 75 individuals were observed using this method. 146 147 Statistical analysis 148 149 Data were recorded as blinks per minute with group size recorded for that minute. 150 Individuals were recorded a mean of 4.9 (± 2.0 SD) minutes before the data collection had to 151 be discontinued due to the observer’s view of the eyes being obstructed. For each individual, 152 we calculated the mean number of blinks in a minute, and the mean group size per minute. 153 To compensate for any unevenness caused by some mean values being based on more 154 observations than others, we conducted the same analyses using just the first minute of data 155 for all individuals (see the Supplementary Material): these data gave qualitatively similar 156 results to the analysis involving mean group sizes, and are not discussed further. 157 158 Using R 4.0.3 [38], we constructed a linear mixed effects model where the natural logarithm 159 of mean blink rate was described by the natural logarithm of group size, and the maturity 160 and sex of the focal individual, including the observation date as a random effect. 161 Logarithms were used to satisfy model conditions. A full model including interactions was 162 initially considered, but no interaction terms were significant and so the basic additive model 163 with the three explanatory variables was used. The Supplementary Material describes 164 resampling analysis conducted to explore any effects of pseudoreplication arising from 165 potentially observing the same individuals multiple times. 166 167 168 169 RESULTS 170

171 Blink rate increased with group size (t67 = 11.38, p < 0.001; figure 1), and adults blinked

172 more than young deer (t67 = 2.11, p = 0.038; figure 1). There was no relationship between

173 sex and blinking (t67 = 0.35, p = 0.727). Model estimates and effect sizes are given in the 174 Supplementary Material, along with analyses demonstrating that pseudoreplication is 175 unlikely to have influenced the group size effect. 176

5 bioRxiv preprint doi: https://doi.org/10.1101/2020.08.11.245837; this version posted January 30, 2021. The copyright holder for this preprint (which was not certified by peer review) is the author/funder, who has granted bioRxiv a license to display the preprint in perpetuity. It is made available under aCC-BY 4.0 International license.

3

2

1 log(blinks per minute)

0 0123 log(group size) 177 178 179 Figure 1. Scatterplot showing that blink rate increases with group size and maturity (where 180 orange triangles and the fitted solid line represent young individuals, and black circles and 181 the fitted dashed line represent adults). 182 183 184 185 DISCUSSION 186 187 Our results demonstrate that blinking increases as group size increases. Given that blinking 188 interferes with vigilance behaviour, and that individual vigilance is predicted to decrease as 189 groups get larger, this supports our argument that the blink rate represents a trade-off 190 between gaining visual information through vigilance and the physiological benefits of 191 blinking. We note that these results are only correlational, and we suggest that a link 192 between vigilance and blinking could be demonstrated with suitable experimental 193 manipulation of perceived risk (e.g. [8,39–41]). 194 195 We argued earlier that observed behaviours that are typically recorded as vigilance (such as 196 holding a head upright or active scanning – see [42,43] for discussion) may not be 197 conducted solely for vigilance. Similarly, blinking may not solely be a maintenance behaviour 198 that is traded-off against being able to collect visual information. Blinking may include a 199 social element, as rhesus macaques Macaca mulatta are able entrain their blink rate in 200 response to social cues, coordinating their blinking with partners that they were interacting 201 with [44]. Our results suggest that the proportion of scanning time that an individual red deer 202 allocates to vigilance is related to the size of its immediate group, but we should 203 acknowledge that scanning behaviour may be influenced by social behaviour as well as

6 bioRxiv preprint doi: https://doi.org/10.1101/2020.08.11.245837; this version posted January 30, 2021. The copyright holder for this preprint (which was not certified by peer review) is the author/funder, who has granted bioRxiv a license to display the preprint in perpetuity. It is made available under aCC-BY 4.0 International license.

204 vigilance. Earlier studies showed that young red deer were less vigilant than older ones 205 [32,33], and suggest this may be because the younger individuals are unlikely to be able to 206 outrun a predator if one appears. If the young in our study are following this behavioural 207 pattern, this also suggests that blinking may not be completely correlated with vigilance 208 level, as the younger deer had low blink rates when compared to mature adults. This may 209 reflect some form of social signalling between adults, although we note that we did not see a 210 difference between males and females (which contradicts previous results showing sex- 211 determined patterns of vigilance, where only females altered behaviour proportionally to 212 group size [32,33]).Other aspects of social behaviour may also be important for determining 213 blink rates, such as position within the group (echoing the vigilance changes seen in [33]). 214 Theory predicts that individuals on the outside of the group should be more vigilant than 215 those in the middle, and anecdotal evidence from olive baboons suggests that peripheral 216 individuals may blink less [31]. It may be possible to test many of the standard predictions 217 connecting vigilance and group size (e.g. [13–16]) using blink rate as a proxy for vigilance. 218 219 The blink rate may also be influenced by factors other than the size of the group, including 220 rainfall and wind (which have been shown to influence blink rate in captive grackles 221 Quiscalus mexicanus [45,46]). Similarly, the behaviour that an individual conducts 222 simultaneously to the blink may be important. Experiments in humans suggest a mechanism 223 controlling the timing of blinks, which occurs to minimise the chance of missing crucial 224 information [20,47–50], with evidence of similar behaviours in rhesus macaques Macaca 225 mulatta [44]. Peafowl Pavo cristatus also time their blinks to coincide with gaze shifts [51] 226 while grackles blink less during flight behaviours [52] and chickens blink more when feeding 227 when compared to scanning [30], all minimising the time where visual information cannot be 228 collected from the environment. Therefore, if an individual is moving, the timing and 229 frequency of its blinks may reflect this movement, and it may therefore be sensible to assay 230 blinking in response to group size in resting deer groups, which would not be undergoing 231 head and body movements that could confound the measure of blinking that is recorded. 232 Similarly, animals in different attentional states may change their frequency of blinking, such 233 as during sleep in herring gulls Larus argentatus [53]. 234 235 Our results suggest that the measurement of blinking presents a simple and non-invasive 236 technique for observing attention that can be conducted remotely. Although we conducted 237 our sampling in the field, this could be done using video footage. Being able to analyse video 238 footage means that information about blink duration can also be collected, and previous 239 studies have demonstrated that this additional metric can also vary between individuals and 240 species [22,23,30,31,45,46], and may increase in relation to group size [30]. Given that

7 bioRxiv preprint doi: https://doi.org/10.1101/2020.08.11.245837; this version posted January 30, 2021. The copyright holder for this preprint (which was not certified by peer review) is the author/funder, who has granted bioRxiv a license to display the preprint in perpetuity. It is made available under aCC-BY 4.0 International license.

241 blinking has been shown to decrease under stressful conditions [28,29], this simple 242 technique could help us to understand the welfare requirements of managed animals that 243 normally live in social groups. 244

8 bioRxiv preprint doi: https://doi.org/10.1101/2020.08.11.245837; this version posted January 30, 2021. The copyright holder for this preprint (which was not certified by peer review) is the author/funder, who has granted bioRxiv a license to display the preprint in perpetuity. It is made available under aCC-BY 4.0 International license.

245 ACKNOWLEDGMENTS 246

247 ZWR and JHR conducted this work as their final year BSc projects. We thank the the School 248 of Biological Sciences teaching lab team for help with equipment, and two anonymous 249 reviewers whose comments greatly improved the manuscript,

250 251 252 FUNDING 253 254 SAR was supported a Natural Environment Research Council (UK) standard grant 255 (NE/P012639/1). 256 257 258 259 REFERENCES 260 261 1. Krause J, Ruxton GD. 2002 Living in groups. Oxford: Oxford University Press.

262 2. Ward A, Webster M. 2016 Sociality: the behaviour of group-living animals. Switzerland: 263 Springer.

264 3. Bednekoff PA, Lima SL. 1998 Re–examining safety in numbers: interactions between risk 265 dilution and collective detection depend upon predator targeting behaviour. Proc R Soc 266 B 265, 2021–2026. (doi:10.1098/rspb.1998.0535)

267 4. Hamilton WD. 1971 Geometry for the selfish herd. J Theor Biol 31, 295–311. 268 (doi:10.1016/0022-5193(71)90189-5)

269 5. Pulliam HR. 1973 On the advantages of flocking. J Theor Biol 38, 419–422. 270 (doi:10.1016/0022-5193(73)90184-7)

271 6. Lima SL. 1990 The influence of models on the interpretation of vigilance. In Interpretation 272 and explanation in the study of animal behavior, volume 2: explanation, evolution and 273 adaptation (eds M Bekoff, D Jamieson), pp. 246–267. Boulder: Westview Press.

274 7. McNamara JM, Houston AI. 1992 Evolutionarily stable levels of vigilance as a function of 275 group size. Anim Behav 43, 641–658.

276 8. Mónus F, Barta Z. 2016 Is foraging time limited during winter? – a feeding experiment 277 with tree sparrows under different predation risks. Ethology 122, 20–29. 278 (doi:10.1111/eth.12439)

279 9. Carere C, Montanino S, Moreschini F, Zoratto F, Chiarotti F, Santucci D, Alleva E. 2009 280 Aerial flocking patterns for wintering starlings, Sturnus vulgaris, under different predation 281 risk. Animal Behaviour 77, 101–107. (doi:10.1016/j.anbehav.2008.08.034)

9 bioRxiv preprint doi: https://doi.org/10.1101/2020.08.11.245837; this version posted January 30, 2021. The copyright holder for this preprint (which was not certified by peer review) is the author/funder, who has granted bioRxiv a license to display the preprint in perpetuity. It is made available under aCC-BY 4.0 International license.

282 10. Rands SA, Cowlishaw G, Pettifor RA, Rowcliffe JM, Johnstone RA. 2003 The 283 spontaneous emergence of leaders and followers in a foraging pair. Nature 423, 432– 284 434. (doi:10.1038/nature01630)

285 11. Rands SA. 2011 Approximating optimal behavioural strategies down to rules-of-thumb: 286 energy reserve changes in pairs of social foragers. PLoS One 6, e22104. 287 (doi:10.1371/journal.pone.0022104)

288 12. Fernández-Juricic E, Smith R, Kacelnik A. 2005 Increasing the costs of conspecific 289 scanning in socially foraging starlings affects vigilance and foraging behaviour. Anim. 290 Behav. 69, 73–81. (doi:10.1016/j.anbehav.2004.01.019)

291 13. Beauchamp G. 2019 On how risk and group size interact to influence vigilance. Biol Rev 292 94, 1918–1934. (doi:10.1111/brv.12540)

293 14. Beauchamp G. 2008 What is the magnitude of the group-size effect on vigilance? Behav 294 Ecol 19, 1361–1368. (doi:10.1093/beheco/arn096)

295 15. Beauchamp G. 2003 Group-size effects on vigilance: a search for mechanisms. Behav 296 Process 63, 111–121.

297 16. Elgar MA. 1989 Predator vigilance and group size in mammals and birds: a critical 298 review of the empirical evidence. Biol Rev 64, 13–33. (doi:10.1111/j.1469- 299 185X.1989.tb00636.x)

300 17. Allan ATL, Hill RA. 2018 What have we been looking at? A call for consistency in studies 301 of primate vigilance. Am J Phys Anthropol 165, 4–22. (doi:10.1002/ajpa.23381)

302 18. Treves A. 2000 Theory and method in studies of vigilance and aggregation. Anim Behav 303 60, 711–722.

304 19. Blount WP. 1927 Studies of the movements of the eyelids of animals: blinking. Q J Exp 305 Physiol 18, 111–125. (doi:10.1113/expphysiol.1927.sp000426)

306 20. Ranti C, Jones W, Klin A, Shultz S. 2020 Blink rate patterns provide a reliable measure 307 of individual engagement with scene content. Sci Rep 10, 8267. (doi:10.1038/s41598- 308 020-64999-x)

309 21. Stevens JR, Livermore A. 1978 Eye blinking and rapid eye movement: pulsed photic 310 stimulation of the brain. Exp Neurology 60, 541–556. (doi:10.1016/0014-4886(78)90009- 311 2)

312 22. Tada H, Omori Y, Hirokawa K, Ohira H, Tomonaga M. 2013 Eye-blink behaviors in 71 313 species of primates. PLoS One 8, e66018. (doi:10.1371/journal.pone.0066018)

314 23. Rands SA. 2021 Phylogenetically-controlled correlates of primate blinking behaviour. 315 PeerJ in press.

316 24. Kirsten SJ, Kirsten EB. 1983 Spontaneous blink rates of birds. Condor 85, 92–93.

317 25. Volkmann FC, Riggs LA, Moore RK. 1980 Eyeblinks and visual suppression. Science 318 207, 900–902. (doi:10.1126/science.7355270)

319 26. Riggs LA, Volkmann FC, Moore RK. 1981 Suppression of the blackout due to blinks. 320 Vision Res 21, 1075–1079. (doi:10.1016/0042-6989(81)90012-2)

10 bioRxiv preprint doi: https://doi.org/10.1101/2020.08.11.245837; this version posted January 30, 2021. The copyright holder for this preprint (which was not certified by peer review) is the author/funder, who has granted bioRxiv a license to display the preprint in perpetuity. It is made available under aCC-BY 4.0 International license.

321 27. Bristow D, Haynes J-D, Sylvester R, Frith CD, Rees G. 2005 Blinking suppresses the 322 neural response to unchanging retinal stimulation. Curr Biol 15, 1296–1300. 323 (doi:10.1016/j.cub.2005.06.025)

324 28. Cross DJ, Marzluff JM, Palmquist I, Minoshima S, Shimizu T, Miyaoka R. 2013 Distinct 325 neural circuits underlie assessment of a diversity of natural dangers by American crows. 326 Proc R Soc B 280, 20131046. (doi:10.1098/rspb.2013.1046)

327 29. Merkies K, Ready C, Farkas L, Hodder A. 2019 Eye blink rates and eyelid twitches as a 328 non-invasive measure of stress in the domestic horse. Animals 9, 562. 329 (doi:10.3390/ani9080562)

330 30. Beauchamp G. 2017 Half-blind to the risk of predation. Front Ecol Evol 5, 131. 331 (doi:10.3389/fevo.2017.00131)

332 31. Matsumoto-Oda A, Okamoto K, Takahashi K, Ohira H. 2018 Group size effects on inter- 333 blink interval as an indicator of antipredator vigilance in wild baboons. Sci Rep 8, 10062. 334 (doi:10.1038/s41598-018-28174-7)

335 32. Childress MJ, Lung MA. 2003 Predation risk, gender and the group size effect: does elk 336 vigilance depend upon the behaviour of conspecifics? Anim Behav 66, 389–398. 337 (doi:10.1006/anbe.2003.2217)

338 33. Lung MA, Childress MJ. 2007 The influence of conspecifics and predation risk on the 339 vigilance of elk (Cervus elaphus) in Yellowstone National Park. Behav Ecol 18, 12–20. 340 (doi:10.1093/beheco/arl066)

341 34. Clutton-Brock TH, Guinness FE, Albon SD. 1982 Red deer: behavior and ecology of two 342 sexes. Chicago: Chicago University Press.

343 35. Rands SA, Muir H, Terry NL. 2014 Red deer synchronise their activity with close 344 neighbours. PeerJ 2, e344. (doi:10.7717/peerj.344)

345 36. Mitchell B, Staines BW, Welch D. 1977 Ecology of red deer: a research review relevant 346 to their management in Scotland. Cambridge: Institute of Terrestrial Ecology.

347 37. Rands SA. 2015 Nearest-neighbour clusters as a novel technique for assessing group 348 associations. Royal Society Open Science 2, 140232. (doi:10.1098/rsos.140232)

349 38. R Development Core Team. 2019 R: a language and environment for statistical 350 computing. Vienna: R Foundation for Statistical Computing.

351 39. Fey K, Banks PB, Ylönen H, Korpimäki E. 2010 Behavioural responses of voles to 352 simulated risk of predation by a native and an alien mustelid: an odour manipulation 353 experiment. Wildl Res 37, 273–282. (doi:10.1071/WR08031)

354 40. Abbey-Lee RN, Mathot KJ, Dingemanse NJ. 2016 Behavioral and morphological 355 responses to perceived predation risk: a field experiment in passerines. Behav Ecol 27, 356 857–864. (doi:10.1093/beheco/arv228)

357 41. Rands SA, Cuthill IC. 2001 Separating the effects of predation risk and interrupted 358 foraging upon mass changes in the blue tit Parus caeruleus. Proceedings of the Royal 359 Society B 268, 1783–1790. (doi:10.1098/rspb.2001.1653)

11 bioRxiv preprint doi: https://doi.org/10.1101/2020.08.11.245837; this version posted January 30, 2021. The copyright holder for this preprint (which was not certified by peer review) is the author/funder, who has granted bioRxiv a license to display the preprint in perpetuity. It is made available under aCC-BY 4.0 International license.

360 42. Mónus F. 2018 Competing activities as measures of fear and vigilance. Anim Sentience 361 2. (doi:10.51291/2377-7478.1312)

362 43. Beauchamp G. 2015 Animal vigilance: monitoring predators and competitors. London: 363 Academic Press. See https://doi.org/10.1016/C2014-0-01423-2.

364 44. Ballesta S, Mosher CP, Szep J, Fischl KD, Gothard KM. 2016 Social determinants of 365 eyeblinks in adult male macaques. Sci Rep 6, 38686. (doi:10.1038/srep38686)

366 45. Yorzinski JL, Argubright S. 2019 Wind increases blinking behavior in great-tailed 367 grackles (Quiscalus mexicanus). Front Ecol Evol 7, 330. (doi:10.3389/fevo.2019.00330)

368 46. Yorzinski JL. 2020 Blinking behavior in great-tailed grackles (Quiscalus mexicanus) 369 increases during simulated rainfall. Ethology 126, 519–527. (doi:10.1111/eth.13003)

370 47. Nakano T, Yamamoto Y, Kitajo K, Takahashi T, Kitazawa S. 2009 Synchronization of 371 spontaneous eyeblinks while viewing video stories. Proc R Soc B 276, 3635–3644. 372 (doi:10.1098/rspb.2009.0828)

373 48. Shin YS, Chang W, Park J, Im C-H, Lee SI, Kim IY, Jang DP. 2015 Correlation between 374 inter-blink interval and episodic encoding during movie watching. PLoS One 10, 375 e0141242. (doi:10.1371/journal.pone.0141242)

376 49. Wiseman RJ, Nakano T. 2016 Blink and you’ll miss it: the role of blinking in the 377 perception of magic tricks. PeerJ 4, e1873. (doi:10.7717/peerj.1873)

378 50. Maffei A, Angrilli A. 2019 Spontaneous blink rate as an index of attention and emotion 379 during film clips viewing. Physiol Behav 204, 256–263. 380 (doi:10.1016/j.physbeh.2019.02.037)

381 51. Yorzinski JL. 2016 Eye blinking in an avian species is associated with gaze shifts. Sci 382 Rep 6, 32471. (doi:10.1038/srep32471)

383 52. Yorzinski JL. 2020 A songbird inhibits blinking behaviour in flight. Biol Lett 16, 20200786. 384 (doi:10.1098/rsbl.2020.0786)

385 53. Amlaner CJ, McFarland DJ. 1981 Sleep in the herring gull (Larus argentatus). Anim 386 Behav 29, 551–556. (doi:10.1016/S0003-3472(81)80118-2)

387

12